While the global discourse on Artificial Intelligence is often dominated by the race for high-end GPUs, a quieter but equally lucrative revolution is taking place within China’s hardware supply chain. Two companies, Yuanjie Semiconductor Technology and Innolight Technology, have emerged as the primary 'money printing machines' of the AI era, with stock prices reaching historic heights. This surge reflects a broader market shift where the 'plumbing' of the internet—the optical modules that convert electrical signals into light—is becoming the most critical bottleneck in the AI compute race.
Yuanjie Technology, led by US-educated scientist Zhang Xingang, has recently claimed the title of the 'New Stock King' on China’s A-share market. Zhang, a Tsinghua graduate with a PhD from the University of Southern California, spent years mastering the complexities of laser chips before returning to Shaanxi to launch his venture. His company focuses on the hardest aspect of the value chain: the optical chips that serve as the heart of transmission modules, a sector traditionally dominated by Western incumbents.
The scale of Yuanjie's ascent is nothing short of spectacular, with its share price ballooning by over 1,500% in a single year. By successfully mass-producing high-end laser chips that meet the rigorous standards of global tech giants, the company has transitioned from a struggling startup to a strategic heavyweight. This success highlights a growing trend of elite Chinese returnees leveraging their international expertise to solve 'chokepoint' technology issues within the domestic ecosystem.
In tandem, Innolight Technology has evolved into a global powerhouse for optical module assembly and manufacturing under the leadership of 76-year-old Wang Weixiu. Originally a traditional motor manufacturer, Wang made a visionary, high-stakes gamble in 2016 by acquiring Suzhou Innolight for 2.8 billion RMB. Once dismissed as an over-leveraged move, that acquisition has propelled the company toward a one-trillion-RMB market capitalization, making Wang one of the wealthiest men in China.
The economic logic driving these valuations is simple: AI computing power is only as good as the speed at which data can move. As tech giants like Nvidia, Microsoft, and Meta scale their data centers, the demand for 800G and 1.6T high-speed optical modules has reached a fever pitch. These components have transitioned from peripheral accessories to essential infrastructure, with high-end margins providing significant tailwinds for manufacturers who can maintain yield and scale.
This phenomenon also signals a new cycle for China’s Growth Enterprise Market, which recently hit an 11-year high. Just as previous cycles were defined by internet services or solar energy, the current era belongs to AI infrastructure. For Chinese investors, the focus has shifted from consumer software to 'hard tech'—the physical components that enable the digital world to function at the speed of light.